Cartographic Expertise for User Maps in Online Geoportals

Abstract

Thanks to the growing availability of geospatial data on online geoportals and mapping platforms, casual map makers are now able to create their own maps easily by combining data layers from different sources. However, the resulting map mashups are often of low cartographic quality. This is due to various and concurring reasons such as combining data originating from different scales and with heterogeneous symbolization, which can lead to serious cartographic issues regarding the content. Furthermore, the visual hierarchy and the symbology of the resulting map are impacted by the absence of cartographic knowledge. To remedy these issues, we define a framework that support the integration of cartographic expertise and functionalities in geoportals in order to improve user-created map mashups. The framework consists of a contextual map model, semantic annotations and formalized cartographic rules. The contextual map model describes the map metadata: its content (data layers), type, scale, and extent, as well as semantic information regarding the layers. Based on the definition of the contextual map model, cartographic principles are turned into formalized rules that can be used to assess and improve the map mashup accordingly. We have developed and tested formalized cartographic rules pertaining to the content of the map, the layer drawing order and the visual hierarchy of the map. The map content is constrained by the map types and associated layer content, while the layer order is based on the geometry types and layer themes (semantics). The function dealing with the visual hierarchy works in two steps. First, the map type and layer selection are analyzed in order to assign each layer to either the background, middle ground, or foreground. Then, the symbolization of the background layers is turned into a style less prominent so as to provide a better figure-ground relationship with the layers in the foreground. The framework offers different options for the background style; two classic options – grayscale and partial desaturation – and a newly developed smart background option. We show whether and how the different options are suitable for background symbolization with concrete map mashups examples. The framework and the associated cartographic rules extend the previously limited cartographic capabilities of the geoportal. The framework is integrated within an existing geoportal using a wizard. The wizard not only helps the user entering information about the map and its purpose, but it is also a platform that allows cartographic knowledge to be shared with the casual map maker via a combination of text and images explanations based on the concrete map example at hand. Finally, the separation of the framework and rules from the implementation makes it possible to transfer the framework to other platform environments. Future areas of research involve the optimal integration of the system within the Graphic User Interface of the geoportal, seconded by the evaluation of interaction types for transmitting cartographic knowledge.